Glucagon‐like peptide‐1 receptor agonists (GLP‐1 RAs) for the management of nonalcoholic fatty liver disease (NAFLD): A systematic review

Abstract There are no licensed drugs for nonalcoholic fatty liver disease (NAFLD), and there is a lack of consensus on the best outcome measures for controlled trials. This systematic review aimed to evaluate the efficacy of GLP‐1 RAs in the management of NAFLD, the degree of heterogeneity in trial design and the robustness of conclusions drawn from these clinical trials. We searched publication databases and clinical trial registries through 2 November 2019 for clinical trials with NAFLD. We evaluated improvements in histological findings, noninvasive markers of hepatic steatosis, inflammation, and fibrosis, insulin resistance and anthropometric measures. Our final analysis included 24 clinical trials, comprising 6313 participants with a mean duration of 37 weeks. Four clinical trials, including RCT (n = 1), single‐arm studies (n = 2) and case series studies (n = 1), used biopsy‐confirmed liver histological change as their end‐points. The remaining studies (n = 20) used surrogate end‐points. GLP‐1 RAs were effective for the improvement in hepatic inflammation, hepatic steatosis and fibrosis. More importantly, GLP‐1 RAs showed promise in improving the histological features of NASH. In addition, 8 ongoing trials were identified. In this systematic review of published and ongoing clinical trials of the efficacy of GLP‐1RAs for NAFLD, we found that GLP‐1 RAs are effective for hepatic steatosis and inflammation, with the potential to reverse fibrosis. Further prospective studies of sufficient duration using histological end‐points are needed to fully assess the efficacy of GLP‐1 RAs in the management of NAFLD.


| INTRODUC TI ON
Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease with a global prevalence of 25.2%, 1 and a higher prevalence of 55.5% in patients with type 2 diabetes mellitus (T2DM). 2 NAFLD is divided into two histological subtypes of (a) nonalcoholic fatty liver (NAFL), characterized by isolated hepatic steatosis, often with mild nonspecific inflammation, and (b) nonalcoholic steatohepatitis (NASH), characterized by the presence of hepatic steatosis and hepatocellular injury with or without fibrosis. NASH is considered to be the more severe form of NAFLD. Approximately 20% of individuals with NASH can progress to cirrhosis, liver failure and hepatocellular carcinoma, while less than 4% of individuals with NAFL progress to cirrhosis. [3][4][5] Patients with T2DM are particularly susceptible to NASH, with a higher risk of progressing into cirrhosis and hepatocellular carcinoma. [6][7][8][9][10] Moreover, the coexistence of NAFLD and T2DM is not only associated with a worse liver outcome but also related to increased risk of extrahepatic diseases, such as cardiovascular disease and chronic kidney disease. 11,12 Therefore, altering the natural course of NAFLD, particularly in T2DM patients, is vital for reducing the health and economic burden of NAFLD and NAFLD-related extrahepatic diseases ( Figure 1).
Lifestyle intervention, the first line of treatment for T2DM and obesity, has proven to be effective in the management of NAFLD.
Reduction of 5%-10% in body weight with life modification over 24-48 weeks leads to a significant improvement in hepatic steatosis, necroinflammation and even fibrosis. [13][14][15][16][17] However, lifestyle intervention alone rarely achieves a complete resolution of NASH and it is challenging to maintain long-term weight loss. Therefore, many pharmacological interventions have been investigated to limit the development and progression of NAFLD, although there are no currently licensed drugs for the treatment of NAFLD. 17,18 Given the close association between NAFLD and T2DM, the effect of antidiabetic medicine for the treatment of NAFLD has attracted substantial scientific attention. [18][19][20][21][22][23] Many clinical trials have suggested the emerging role of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) in the management of NAFLD. However, one of the biggest challenges in designing and implementing controlled trials in NAFLD is the lack of consensus on appropriate end-points for assessing the benefit of GLP-1 RAs for NAFLD. 24,25 Although end-points for NAFLD in clinical trials have evolved during the past decades, liver biopsy is still the gold standard for diagnosis and assessment of NAFLD. However, the invasive nature of liver biopsy and reluctance from patients limits its use in clinical trials and thus constitutes a major barrier for drug development in NAFLD. As a result, several noninvasive serum markers or imaging modalities for diagnosis or assessing response to treatment for NAFLD have been developed, and they have been increasingly used for defining endpoints in clinical trials. [26][27][28][29][30][31] Our systematic review aimed to evaluate the efficacy of currently available GLP-1 RAs (Table 1) in the management of NAFLD, F I G U R E 1 Natural history, risk factors and treatment approaches of NAFLD the degree of heterogeneity in trial design and the robustness of conclusions drawn from these clinical trials.

| Data sources and extraction
This systematic review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 32 We conducted a systematic litera- Differences were resolved by discussion with SYQ.

| Selection of published studies
The inclusion criteria were published clinical trials investigating the effect of GLP-1 RAs on NAFLD. The diagnosis of NAFLD was based on the detection of steatosis either by imaging or by histology, and appropriate exclusion of other liver diseases. 33 The exclusion criteria were studies not written in English and those with secondary causes of hepatic steatosis. Reviews and editorials were excluded.
There were no restrictions on sex, age, ethnicity and numbers of participants.

| Selection of ongoing registered clinical trials
ClinicalTrials.gov was searched to identify ongoing registered clinical trials. The inclusion and exclusion criteria were the same as those for the selection of published studies.

| Outcome measures
The primary outcome assessed in clinical trials included histological improvement in NAFLD, defined as the resolution of steatohepatitis without worsening of fibrosis. Secondary histological outcomes included steatosis, hepatocyte ballooning, (lobular or portal) inflammation and the combined NAFLD activity score. Liver/spleen ratio (CT) improved from 0.92 ± 0.30 to 1.04 ± 0.24, P < .01 Histological inflammation improved in 7 of the 10, liver fibrosis improved in 6 of the 10, and NAFLD activity score improved in 8 of the 10.
14.8% dropout Lira has a good safety profile.

Tolerability Comments Liver fat by imaging Histology
Ins was associated with a significant decrease in liver mean MRI-PDFF (13.8% to 10.6%, P = .005). Lira did not change MRS-PDFF (P = .80).
NA 4 of Lira discontinued due to adverse effects.
Weight was improved (−2.8 ± 6.5 in Lira vs. 0 in Ins, P = .03) NA Lira was associated with significantly increased odds of resolution of definite NASH and progression of fibrosis than placebo group.
13.46% dropout Most adverse events were mild to moderate in severity, transient and similar between groups.
There is no significant improvement in hepatic steatosis ( 1 H-MRS) across three groups. Exe induced a significant reduction in LFC ( 1 H-MRS) in the Exe group than in the PLA group (−23.8 ± 9.5% vs + 12.5 ± 9.6%, P = .007) NA 13.6% dropout Longer exposure time to exenatide might be needed to reveal significant improvement in myocardial triglyceride content.
Liver steatosis (CAP) was not improved.
Only one case had a liver biopsy.
The total NAFLD activity score was improved from 6 to 2.

NA No dropout
Both groups had significant reductions in liver stiffness (P = .003). No significant difference existed between groups.

15.0% dropout
The effect of Lira in reducing LFC was mainly driven by bodyweight reduction.
LFC was significantly reduced in all groups, from 36

| Quality assessment
The quality of randomized control trials (RCTs) was assessed based on a modified version of the Cochrane Collaboration Risk of Bias Tool. 34

| Study design and selection of end-points
Four of 24 studies, including RCT (n = 1), single-arm studies (n = 2) and case series studies (n = 1), used biopsy-confirmed liver histological change as their end-point. The remaining studies (n = 20) used surrogate end-points, including change in hepatic enzymes (n = 21), noninvasive assessment of hepatic steatosis (n = 8) and liver fibrosis (n = 9). on the intrahepatic lipid (IHL) measured by 1 H MRS. 65 In this study, GLP-1 RA treatment was associated with a 42% relative reduction in IHL (−59.3, −16.5%) (P < .01), and the most considerable IHL reduction occurred among patients with highest pretreatment levels.

| Study interventions
Likewise, Dutour et al, in a prospective randomized trial enrolling a total of 44 obese subjects with T2DM randomly assigned to receive exenatide or reference treatment, found a substantial reduction in liver fat content in the exenatide group (−23.8 ± 9.5%) versus the reference group (+12.5 ± 9.6%) (P = .007). 63 Participants in the exenatide group also had a more significant reduction in insulin resis- Liver fat content was significantly reduced with pioglitazone treatment (11.0 ± 3.1 to 6.5 ± 1.9%, P < .05), and combined pioglitazone and exenatide therapy was linked to a more significant decrease in hepatic fat (12.1 ± 1.7 to 4.7 ± 1.3%, P < .05

| GLP-1 RAs for the treatment of hepatic fibrosis
The primary objective of treatment for NASH is to prevent the development of cirrhosis, and increasing hepatic fibrosis is the hallmark of disease progression to cirrhosis. 50   One of 8 will use the number of treatment-emergent adverse events, serious adverse events and any grade ≥ 1 laboratory abnormality as the primary end-point. Of 7 controlled trials, 3 plan to use placebo, 2 use lifestyle intervention, and 2 use antidiabetic medicines or other medicine targeting NASH as their control. On review of ongoing trials, there has been a general shift from the use of hepatic enzymes and ultrasonographical findings as end-points to MRI assessment of hepatic steatosis, liver fibrosis and biopsy-confirmed liver histology.

| D ISCUSS I ON
Early evidence of GLP-1 RAs for NAFLD comes from studies reporting an improvement in hepatic enzymes with exenatide therapy. 59,85 These  resistance. 41 Adiponectin is associated positively with insulin sensitivity, promoting fatty acid β-oxidation (FAO), glucose use and suppression of fatty acid synthesis. 42,43 Patients with NAFLD have a lower level of adiponectin compared with BMI-matched controls. 44 Collectively, NAFLD is closely associated with both hepatic and adipose tissue insulin resistance, and reduced systemic insulin resistance ( Figure 2). 45 This review has highlighted the limitations of the current data for the treatment of NAFLD. A major issue that hinders drug development for NAFLD is the need for biopsy-confirmed liver histology to evaluate the severity of disease and assess response to therapies.
It is critical to measure disease severity, particularly the presence of NASH and the stage of fibrosis because NASH and fibrosis severity have been strongly implicated in the long-term prognosis of NAFLD. Although liver biopsy, in combination with Kleiner's histological NAFLD activity score (NAS), is still the gold standard for the stage of NASH, it is impractical to perform a liver biopsy in a large sample group because of the potential risk of infection and bleeding.
Another limitation of liver biopsy is that the volume of a needle bi-

ACK N OWLED G EM ENTS
This study is funded by the Health Commission of Jilin Province, China (Grant No.20152019).

CO N FLI C T O F I NTE R E S T
The authors have declared no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data related to this study are included in the manuscript and supplementary materials.